Supplemental material to this article can be found at:
http://molpharm.aspetjournals.org/content/suppl/2012/10/05/mol.112.081489.DC1
1521-0111/83/1/73–84$25.00
MOLECULAR PHARMACOLOGY
Copyright ª 2013 by The American Society for Pharmacology and Experimental Therapeutics
http://dx.doi.org/10.1124/mol.112.081489
Mol Pharmacol 83:73–84, January 2013
Salt Bridge Switching from Arg290/Glu167 to Arg290/ATP
Promotes the Closed-to-Open Transition of the P2X2 Receptor s
Ralf Hausmann, Janka Günther, Achim Kless, Daniel Kuhlmann, Matthias U. Kassack,
Gregor Bahrenberg, Fritz Markwardt, and Günther Schmalzing
Molecular Pharmacology, RWTH Aachen University, Aachen, Germany (R.H., J.G., D.K., G.S.); Grünenthal GmbH, Global Drug
Discovery, Departments of Molecular Pharmacology and Discovery Informatics, Aachen, Germany (A.K., G.B.); Institute of
Pharmaceutical and Medicinal Chemistry, Heinrich Heine University, Düsseldorf, Germany (M.U.K.); and Julius Bernstein Institute
for Physiology, Martin Luther University, Halle/Saale, Germany (F.M.)
ABSTRACT
P2X receptors are trimeric adenosine-5’-triphosphate (ATP)gated cation channels involved in fast signal transduction in
many cell types. In this study, we used homology modeling of
the rat P2X2 receptor with the zebrafish P2X4 X-ray template to
determine that the side chains of the Glu167 and Arg290
residues are in close spatial vicinity within the ATP-binding
pocket when the rat P2X2 channel is closed. Through charge
reversal mutation analysis and mutant cycle analysis, we
obtained evidence that Glu167 and Arg290 form an electrostatic
interaction. In addition, disulfide trapping indicated the close
proximity of Glu167 and Arg290 when the channel is in the
closed state, but not in the ATP-bound open state. Consistent
Introduction
P2X receptors are adenosine-5’-triphosphate (ATP)-gated
cation channels assembled from a repertoire of seven
homologous subunits (P2X1 though P2X7) into homotrimers and heterotrimers (Nicke et al., 1999; North, 2002;
Kaczmarek-Hajek et al., 2012). A crucial advance in the
understanding of the P2X receptor at the molecular level is
the resolution of the X-ray crystal structure of the zebrafish
P2X4.1 (zfP2X4) receptor in the apo-closed and the ATPbound open state (Kawate et al., 2009; Hattori and Gouaux,
2012). The zfP2X4 protomer structure has a “dolphin” shape
(Kawate et al., 2009) (see Fig. 1B) in which the intersubunit
ATP-binding pocket is formed by the head domain, the upper
body, and the left flipper of one subunit and the lower body
and dorsal fin of an adjacent subunit. In the X-ray structure of
the ATP-bound zfP2X4 receptor, the ATP molecule is accommodated into the binding pocket in a U-shaped configuration
in which the b- and g-phosphates are positioned toward the
This work was supported by the German Research Foundation (Grants
Schm 536/8-1, Schm 536/8-2).
dx.doi.org/10.1124/mol.112.081489.
s This article has supplemental material available at molpharm.
aspetjournals.org.
with a gating-induced movement that disrupts the Glu167/
Arg290 salt bridge, a comparison of the closed and open rat
P2X2 receptor models revealed a significant rearrangement of
the protein backbone and the side chains of the Glu167 and
Arg290 residues during the closed-to-open transition. The
associated release of the Glu167/Arg290 salt bridge during
channel opening allows a strong ionic interaction between
Arg290 and a g-phosphate oxygen of ATP. We conclude from
these results that the state-dependent salt bridge switching
from Arg290/Glu167 to Arg290/ATP fulfills a dual role: to
destabilize the closed state of the receptor and to promote the
ionic coordination of ATP in the ATP-binding pocket.
adenine ring (Hattori and Gouaux, 2012). This conformation
allows the formation of a series of salt bridges and hydrogen
bonds between the acidic phosphate oxygen atoms of ATP
and a cluster of highly conserved basic and polar residues,
including Lys70 and Lys72 of one subunit and Asn296, Arg298,
and Lys316 of the adjacent subunit (Hattori and Gouaux,
2012). The corresponding residues have also been suggested to
be involved in ATP binding in other P2X receptor subtypes
based on the results of mutagenesis studies and computational
homology modeling (Guerlet et al., 2008; Roberts et al., 2008,
2012; Ennion et al., 2000; Roberts and Evans, 2004; Fischer
et al., 2007; Allsopp et al., 2011; Du et al., 2012).
Normal mode analysis and data from P2X2 mutants with
introduced histidines, which allow Zn2+-mediated domain
bridging, have showed that ATP binding induces the head
domain and dorsal fin to move closer to each other, which
leads to a tightening of the ATP-binding pocket (Jiang et al.,
2012). This conformational closure of the intersubunit cleft
between the head and dorsal fin domains was also observed in
a comparison of the X-ray structures of the apo- and ATPbound states of the zfP2X4 receptor (Hattori and Gouaux,
2012). The extracellular vestibule of the ATP-binding pocket
is simultaneously enlarged by the outward flexion and rotation
ABBREVIATIONS: ATP, adenosine-5’-triphosphate; Cy5 NHS ester, Cy5 N-hydroxysuccinimide ester; DTT, dithiothreitol; IC50, 50% inhibitory
concentration; ab-meATP, ab-methylene-ATP; MTS, methanethiosulfonate; MTSEA, 2-aminoethyl methanethiosulfonate; MTSES, 2-sulfonatoethyl
methanethiosulfonate; NF770, 7,7’-(carbonylbis(imino-3,1-phenylenecarbonylimino-3,1-(4-methyl-phenylene)carbonylimino))bis(1-methoxy-naphthalene-3,6-disulfonic acid); PDB, Protein Data Bank; TEVC, two-electrode voltage-clamp; zf, zebrafish.
73
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
Received July 23, 2012; accepted October 5, 2012
74
Hausmann et al.
of the lower body domains. This movement is transmitted to
the directly linked transmembrane domains TM1 and TM2 to
induce an iris-like movement of the TM helices, which
ultimately opens the ion channel pore (Hattori and Gouaux,
2012). Additional evidence on the rotational conformational
changes of TM2, which underlie the pore opening of the P2X
receptors and allow ion access to the pore through lateral
portals adjacent to the membrane, has also been obtained using
mutagenesis data (Li et al., 2010; Cao et al., 2009; Kracun et al.,
2010; Samways et al., 2011; Kawate et al., 2011). An extensive
rearrangement within the extracellular domain, including the
pulling apart of the b-strands of adjacent subunits, was
further suggested by high-resolution electron microscopy
and computational studies on the hP2X1 and zfP2X4 receptor,
respectively (Du et al., 2012; Roberts et al., 2012).
We previously used the X-ray structure of the closed zfP2X4
receptor (Kawate et al., 2009) (PDB entry 3H9V) as a template
to generate a homology model of the rP2X2 receptor. This
model allowed us to identify the molecular determinants of
the nanomolar potency interaction of the suramin-derivative
NF770 with the P2X2 receptor (Wolf et al., 2011). The model
predicted critical roles for Glu167 and Arg290 in ATP and
NF770 binding, and these predictions were experimentally
verified. In the present study, we investigated whether the
close proximity of Glu167 and Arg290, which is predicted by
our closed-state P2X2 model (Fig. 1A), allows an electrostatic
interaction that may be important for receptor function. The
residues Glu167 and Arg290 are conserved in all P2X receptor
subtypes across several species, which indicates their potential importance in P2X receptor function. We provide evidence
that a Glu167/Arg290 salt bridge is formed when the P2X2
receptor closes. This salt bridge is released when Glu167 and
Arg290 move away from each other upon ATP binding and
channel opening, which permits a strong ionic interaction between Arg290 and a g-phosphate oxygen of ATP. By restricting
the free movement of Glu167 and Arg290, this salt bridge
switching may help guide the conformational transitions that
define the closed and ATP-bound open states of the channel.
Materials and Methods
Chemicals. The methanethiosulfonate (MTS) compounds 2-aminoethyl methanethiosulfonate (MTSEA) and 2-sulfonatoethyl methanethiosulfonate (MTSES) were purchased from Biotium (Hayward, CA).
ATP (sodium salt) was purchased from Roche (Roche Diagnostics;
Mannheim, Germany). The Cy5 NHS ester was purchased from GE
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
Fig. 1. Homology model of the closed state of the rP2X2 receptor. (A) the homotrimeric rP2X2 receptor is shown from the side in a partially transparent
surface representation in which the subunits are colored light gray, purple, and brown. One of the three ATP-binding pockets is marked by a rectangular
frame and is shown in a close-up view using a stick representation of the side chains of Glu167 and Arg290. The specific atoms of Glu167 and Arg290 are
colored by element type: carbons are cyan or pink, hydrogens are gray, oxygens are red, and nitrogens are blue. The carboxylate carbon of Glu167 is 3.9 Å
distant from the guanidinium carbon of Arg290. The side chains of Lys69 and Lys308, which are also important for ATP binding, are also shown as sticks
(carbons are white, hydrogens are gray, and nitrogens are blue). The brown-colored subunit is shown in the ribbon representation, and the coloring of the
protein domains is the same as that in Fig. 1B. (B) one rP2X2 subunit in the closed state is shown from the side through a ribbon representation in which
the a-helices and b-strands are consecutively numbered. The gray contours illustrate the dolphin-like shape of this single subunit (the body is blue, the
fluke is green, the head is pink, the dorsal fin is orange, the right flipper is red, and the left flipper is yellow) (Kawate et al., 2009). The side chains of
Glu167 and Arg290, which are located C-terminal to the b9-strand and within the b16-strand, respectively, are highlighted through a space-filling
representation.
Salt Bridge Switching in P2X2 Receptor Gating
I
top 2 bottom
þ bottom;
¼
Imax
EC50 nH
1þ
½ATP
(1)
where I is the response evoked by the ATP concentration ([ATP]), Imax
is the maximal response, and nH is the Hill coefficient. The bottom
level was constrained to zero. Each oocyte was challenged with seven
increasing concentrations of ATP ranging from 1 mM to 1 mM or from
3 mM to 3 mM in logarithmically equal steps. The EC50 values are
presented as the geometric means and the corresponding 95%
confidence intervals (95% CI). The mean current amplitudes and Hill
slopes are presented as the arithmetic mean 6 S.E.M. Error bars were
omitted in the figures when these were smaller than the symbols used.
The effect of MTS or H2O2 treatment on the mean current amplitudes
mediated by various P2X2 constructs was calculated as follows: (Iafter/
Iinitial) 100, where Iinitial and Iafter are the averaged ATP current
amplitudes measured before and after the indicated treatment,
respectively. The effect of the chemical reduction by DTT after H2O2
treatment was calculated as (Iafter H2O2/DTT/Iafter H2O2) 100, where
IH2O2 and IH2O2/DTT represent the current amplitudes recorded before
and after the application of DTT, respectively. Provided that a
normal distribution could be verified by the D’Agostino-Pearson
omnibus K2 normality test, one-way analysis of variance (ANOVA)
followed by a post hoc Bonferroni’s multiple comparison test was
used to determine the level of statistical significance between the
means, as indicated in the figure legends.
The change in the Gibbs free energy (DDG) for each mutant (mut)
was calculated as
DDG ¼ 2 RT×ln
!
EC50 wt
;
EC50 mut
(2)
where EC50wt and EC50mut represent the EC50 values of the wt and
mutant rP2X2 receptors, respectively. The constants R and T have the
values of 1.987 cal/mol/K and 293 K, respectively. The interaction free
energy, which is also designated as the coupling energy of the interaction of two residues (DDGINT), was calculated as
DDGINT ¼ DDGmut1þ2 2 DDGmut1 þ DDGmut2
"
Eq:2
⇒ DDG ¼ 2 RT× ln
EC50 wt
EC50 mut1þ2
2 ln
⇔DDG ¼ 2 RT×ln
!
EC50 wt
EC50 mut1
!
þ ln
EC50 mut1 ×EC50 mut2
EC50 mut1þ2 ×EC50 wt
!
;
EC50 wt
!!#
EC50 mut2
ð3Þ
where EC50mut1 and EC50mut2 represent the EC50 values of the single
mutants and EC50mut1+2 represents the EC50 value of the corresponding double mutant. The experimental error in the calculated
DDG and DDGINT values was determined using the lower and upper
limits of the 95% confidence interval of the EC50 values of the wt
rP2X2 receptor and the respective mutants (Schreiber and Fersht,
1995).
zfP2X4 Structure-Based Homology Models of the rP2X2
Receptor. The detailed procedure for generating the homology
models using the molecular modeling program MOE2008.10 (Molecular Operating Environment 2008; CCG, Montreal, Canada) has been
previously described elsewhere (Wolf et al., 2011). To generate the
closed- and open-state models of the rP2X2 receptor, we used as
templates the X-ray structures of the zfP2X4 receptor in the apoclosed state (PDB entry 3H9V) (Kawate et al., 2009) and the ATPbound open state (PDB entry 4DW1) (Hattori and Gouaux, 2012),
respectively. The final sequence alignment used for the homology
modeling is shown in Supplemental Fig. 1. A comparison of the 3H9Vbased closed-state rP2X2 model with an identically generated closedstate rP2X2 model based on the recently refined PDB 4DW0 structure
(Hattori and Gouaux, 2012) revealed that the orientations and side
chain directions of Glu167 and Arg290 were the same in both templates. For the sake of consistency, we used the 3H9V-based rP2X2
homology model (Wolf et al., 2011) in the present study. The PDB files
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
Healthcare (München, Germany). All other chemicals were obtained
from either Sigma-Aldrich (Taufkirchen, Germany) or Merck (Darmstadt, Germany).
P2X Receptor Expression in Xenopus laevis Oocytes. The
oocyte expression plasmid encoding the N-terminal His6-tagged rat
P2X2 subunit was available from previous studies (Aschrafi et al., 2004;
Hausmann et al., 2006; Meis et al., 2010; Wolf et al., 2011; Franklin
et al., 2007). The mutations were introduced using the QuikChange
site-directed mutagenesis kit (Stratagene, La Jolla, CA). All of the
constructs were verified using restriction analysis and nucleotide
sequencing. The capped cRNAs were synthesized as previously described
elsewhere (Schmalzing et al., 1991; Nicke et al., 1998) and injected
into collagenase-defolliculated X. laevis oocytes in 50-nl aliquots (0.1
mg/ml) using a Nanoliter 2000 injector (WPI, Sarasota, FL). The
oocytes were cultured at 19°C in sterile oocyte Ringer’s solution (ORi:
90 mM NaCl, 1 mM KCl, 1 mM CaCl2, 1 mM MgCl2, and 10 mM
HEPES, pH 7.4) supplemented with 50 mg/ml gentamicin.
Functional Analysis and Modification of Introduced
Cysteines. One to three days after the cRNA injection, the current
responses were evoked through ATP at ambient temperature (21–24°C)
in a nominally calcium-free ORi solution (designated Mg-ORi: 90 mM
NaCl, 1 mM KCl, 2 mM MgCl2, and 10 mM HEPES, pH 7.4) and
recorded using a conventional two-electrode voltage-clamp (TEVC)
with a Turbo TEC-05 amplifier (NPI Electronics, Tamm, Germany)
at a holding potential of 260 mV, as previously described elsewhere
(Rettinger and Schmalzing, 2003; Hausmann et al., 2006). The
cysteine reactivity experiments with oxidizing or reducing agents
were performed using KCl agar bridge bath electrodes. In the ATP
concentration–current response analysis, each P2X2 receptorexpressing oocyte was challenged in 30-second intervals with 15second pulses of increasing concentrations of ATP.
To determine the effect of the MTS reagents, the oocytes expressing
the desired rP2X2 construct were repeatedly stimulated with ATP
until a stable current response was obtained. The ATP-evoked current
amplitudes were recorded before and after a 120-second pulse of 1 mM
MTSEA or 1 mM MTSES followed by a 120-second wash with Mg-ORi.
To determine the effect of MTS on the ATP-bound P2X2 receptor, 3 mM
ATP was applied for 3 seconds to open the P2X2 receptor; this step was
followed by the coapplication of 3 mM ATP with either 1 mM MTSEA or
1 mM MTSES.
To indirectly monitor the disulfide bond formation, a minimum of
two ATP-evoked current amplitudes were recorded before and after the
application of a 120-second pulse of 0.3% H2O2, which was followed by
a 150-second wash with Mg-ORi. To examine the effect of H2O2 on the
open P2X2 receptor, 3 mM ATP was applied for 3 seconds to open the
receptor, followed by the coapplication of 3 mM ATP and 0.3% H2O2.
The reversibility of the effects of H2O2 were tested by recording
a minimum of two ATP-evoked current amplitudes from H2O2-treated
oocytes before and after a 180-second application of 10 mM dithiothreitol
(DTT), which was followed by a 150-second wash with Mg-ORi.
Assembly and Plasma Membrane Expression of the P2X2
Receptor. The assembly and plasma membrane expression of the
various P2X2 receptor constructs were analyzed using blue nativePAGE (BN-PAGE) and SDS-PAGE, as previously described elsewhere
(Hausmann et al., 2012; Fallah et al., 2011).
Data Analysis. The data were plotted and fitted using Prism5
(GraphPad Software Inc., San Diego, CA). The agonist concentration–
response curves and EC50 values were obtained by iteratively fitting
the four-parameter Hill equation to the pooled data points from n
oocytes:
75
76
Hausmann et al.
of the rP2X2 homology modeling data shown in Figs. 1 and 8 are
available in the supplemental data.
Results
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
The Charge-Reversal and Charge-Swap Mutants
Suggest That Glu167 and Arg290 Form an Electrostatic
Interaction. Our rP2X2 homology model (Wolf et al., 2011)
predicts that the carboxylate carbon of Glu167 and the
guanidinium carbon of Arg290 are separated by only ∼3.9 Å
(Fig. 1A). Glu167 is located at the base of the cysteine-rich
head domain directly C-terminal to the b9-strand, which
connects the body domain and the cysteine-rich head domain
in the upper region of the ATP-binding pocket (Fig. 1B).
Arg290 is located in the b16-strand of the body domain at the
base of the left flipper (Fig. 1B) and is part of the conserved
“NFR” motif, which is known to be crucial for the activation of
P2X receptors by ATP. To test for a possible electrostatic
interaction between Glu167 and Arg290, we individually
reversed the charges and swapped the charges to break and
restore the ionic interactions, respectively. The charge reversal of Glu167 with Arg (E167R) resulted in a slight
increase in the ATP potency and a 1.5-fold decrease of the
mean current amplitude elicited by 1 mM ATP compared with
the wt rP2X2 receptor (Fig. 2, A and B; Table 1). The charge
reversal of Arg290 by Glu (R290E) decreased the ATP potency
by a factor of 624 and decreased the mean current amplitude
elicited by 1 mM ATP by 500-fold compared with the wt rP2X2
receptor (Fig. 2, A and B; Table 1). The pronounced effect of
the charge reversal at Arg290 might be due in part to the
crucial role of Arg290 in ATP binding (Roberts et al., 2008;
Hattori and Gouaux, 2012). Biochemical analyses showed
that the homotrimeric subunit assembly and the plasma
membrane expression of the receptor were not affected by any
of the charge-reversal mutations (Fig. 2C).
If the effects of the charge reversals at Glu167 and Arg290
are independent, the effect of the double reversal should be
additive, that is, a total 374-fold decrease in the ATP potency
and a total 714-fold decrease in the mean current amplitude.
However, the charge-swapped mutant E167R,R290E–rP2X2
mutant exhibited a 29- and 2.5-fold decrease in the ATP
potency and the mean current amplitude elicited by 1 mM
ATP, respectively, compared with the wt rP2X2 (Fig. 2, A and
B; Table 1). Thus, the charge swapping rescued the ATP
potency by 12.9-fold compared with the effects of the single
mutants. A similar 3.9- to 12.2-fold rescue of the ATP potency
was observed in all possible charge-swap mutants that can be
generated by exchanging Glu167 with Lys or Arg and Arg290
with Glu or Asp. These data are summarized in Table 1.
The Double-Mutant Cycle Analysis Supports an
Electrostatic Interaction between Glu167 and Arg290.
The functional rescue of the charge-swapped mutants
compared with the single-charge reversal suggests that
Glu167 and Arg290 form an electrostatic interaction with
each other. To further investigate this interaction, we
performed double-mutant cycle analysis, which consists of
the calculation of the interaction energy between a pair of
residues based on the Gibbs free energy change associated
with the modification or mutation of these residues (Horovitz,
1996; Hidalgo and Mackinnon, 1995; Serrano et al., 1990;
Carter et al., 1984; Schreiber and Fersht, 1995). If the
residues do not interact, the change in the free energy of the
double mutant should equal the sum of the free energy
changes of the corresponding single mutations. In contrast, if
the two residues interact, the change in the free energy of the
double mutant will differ from the sum of the changes of
the corresponding single mutations. Using the changes in the
EC50 value, we determined that the additive sum of the
changes in the free energy of the single mutants was different
from the change in the free energy of the E167R,R290E-rP2X2
double mutant (Table 1). The calculated interaction energy
of DDGINT of 21.45 6 0.03 kcal/mol (Fig. 3) was significantly
larger than the threshold value of 60.35 kcal/mol that has
been reported for noninteracting residues (Schreiber and
Fersht, 1995). In general, a negative Gibbs free energy change
due to an amino acid mutation indicates that the interacting
residues have a stabilizing effect on the structure, whereas
a positive free energy change indicates a destabilizing effect
(Kumar and Nussinov, 1999). Thus, the interaction free energy
DDGINT of 21.45 that was calculated from the experimentally
determined EC50 values suggests that the interaction of Glu167
and Arg290 has a stabilizing effect.
In addition, we used the E167A and/or R290A mutants for
double-mutant cycle analysis. The rationale for this experiment was that alanine replacement mutations are less likely
to result in new interactions and are thus considered to be
superior to other mutations in mutant cycle analysis (Faiman
and Horovitz, 1996). The alanine substitution of Glu167 did
not affect the ATP potency, whereas the alanine substitution
of Arg290 decreased the ATP potency by more than 230-fold
(Table 1). An interaction free energy (DDGINT) of 21.11 6 0.04
kcal/mol, which was also above the threshold of 60.35 kcal/
mol for noninteracting residues (Schreiber and Fersht, 1995),
was calculated for the alanine double mutant E167A,R290ArP2X2.
As negative controls, we mutated the oppositely charged
residues Asp57 and Arg275 of the rP2X2 receptor, which are
separated by more than 30 Å according to our homology
model. Consistent with their noninteraction, the ATP potency
of the charge-swapped D57R,R275D-rP2X2 double mutant
differed by only ∼1.6-fold from the additive sum of the ATP
potencies of the charge-reversal mutants D57R-rP2X2 and
R275D-rP2X2. The double-mutant cycle analysis yielded
a low interaction energy (DDGINT) of 0.32 6 0.07 kcal/mol
(Fig. 3; Table 1). In addition, the alanine mutants of Asp57
and Arg275 yielded a nonsignificant interaction energy
(DDGINT) of 20.31 6 0.04 kcal/mol, which is below the 6
0.35 kcal/mol calculated threshold energy (Schreiber and
Fersht, 1995) (Table 1).
The Modification of the Introduced Cysteines with
Charged MTS Reagents Supports a Functionally Important Electrostatic Interaction between Residues
167 and 290. To provide further evidence on the electrostatic
nature of the interaction between Glu167 and Arg290, we
substituted Glu167 and Arg290 individually with cysteine
residues. As a background for these mutations, we used an
rP2X2 mutant (designated rP2X2C9,348,430S) in which the
three cysteine residues, which are located in the cytoplasmic
N-terminal domain (Cys9), the C-terminal end of TM2
(Cys348), and the C-terminal domain (Cys430), were replaced
by serines. The three Cys-to-Ser mutations were functionally
silent in terms of ATP potency and mean current amplitudes
(Fig. 2A; Table 1), but prevented the spontaneous intersubunit
cross-linking that is observed in wt rP2X2 (Jiang et al., 2010;
Salt Bridge Switching in P2X2 Receptor Gating
77
Fig. 2. Functional and biochemical characterization of the Glu167 and
Arg290 rP2X2 mutants. (A) Bar graph showing the mean current
amplitudes (6 S.E.M.) elicited by 1 mM ATP in oocytes expressing the
indicated wt or mutant rP2X2 receptor. The results for the cysteine
mutants, in which cysteines were introduced into the rPX2C9,348,430S
receptor background (lacks nonextracellular cysteine residues), are shown
as gray bars. (B) ATP concentration–response curves obtained with the
indicated wt or mutant rP2X2 receptors. The absolute current amplitudes
are indicated. j, wt-rP2X2, EC50 = 11.4 (95% CI 9.3–14.9) mM, nH = 1.5 6
0.4, n = 9; d, E167R-rP2X2, EC50 = 6.3 (95% CI 5.4–7.6) mM, nH = 1.8 6 0.2,
n = 6; ., R290E-rP2X2, EC50 = 7192 (95% CI 6393–8607) mM, nH = 2.6 6
0.3, n = 6; m, E167R,R290E-rP2X2, EC50 = 328 (95% CI 288–371) mM, nH =
2.6 6 0.2, n = 6. (C) The indicated charge-reversal mutants show a wt-like
homotrimeric assembly and cell surface abundance. The proteins were
purified under nondenaturing conditions from X. laevis oocytes using NiNTA chromatography, resolved by BN-PAGE (upper panel) or reducing
SDS-PAGE (lower panel), and visualized using Typhoon fluorescence
scanning. “SDS” in lane 1 indicates the migration of the rP2X2 protein
after partial denaturation after a 1-hour incubation with 0.1% SDS before
BN-PAGE. The open circles in the left margin indicate the numbers of
rP2X2 subunits that correspond to each protein band. (D) The indicated
proteins were purified as in C, resolved using nonreducing SDS-PAGE and
visualized using Typhoon fluorescence scanning. The rPX2C9,348,430S
receptor and the indicated cysteine mutants are efficiently expressed at
the cell surface and do not exhibit intersubunit cross-linking. The asterisk
indicates a nonspecific background band that was present in some
samples. The numbers in the left margins in C and D refer to the
molecular masses of the marker proteins (in kiloDaltons).
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
Li et al., 2008). Nonreducing SDS-PAGE analysis verified that
the rP2X2C9,348,430S receptor did not form any intersubunit
cross-links and showed efficient plasma membrane expression
of the rP2X2C9,348,430S receptor and the single and double
E167C and R290C mutants (Fig. 2D). The E167C mutation
had no effect on the ATP potency and decreased the mean
current amplitude by a factor of 2. The R290C substitution
resulted in a 71-fold decrease in the ATP potency and a 10-fold
decrease in the mean current amplitude compared with the
parental rP2X2C9,348,430S receptor (Fig. 2A; Table 1). The
strong effect of the R290C mutation might be due to the
crucial role of Arg290 in ATP binding (Hattori and Gouaux,
2012).
To reintroduce a charge at positions 167 and 290 posttranslationally, we used the positively and negatively charged
MTS reagents MTSEA and MTSES, respectively; these reagents share a similar molecular mass and thiol reactivity.
Accordingly, the differences in the current responses to ATP
after treatment with MTSEA or MTSES can be attributed to
the addition of a positive or negative charge at the introduced cysteine, respectively. The treatment of the parental
rP2X2C9,348,430S receptor with 1 mM MTSEA or MTSES for
120 seconds had almost no effect on the current amplitudes
(Fig. 4, left panel; Table 2) or the ATP potency (Table 2). The
restoration of the original charges at E167C or R290C by
MTSES or MTSEA, respectively, resulted in a partial rescue
of the receptor function, which was observed as a 2-fold decrease
in the EC50 (both mutants, Table 2) and a 2.8-fold increase in
the current amplitude (R290C mutant; Fig. 4, right panel; Table
2). In contrast, the introduction of an opposite charge at each
position further decreased the function of the receptor, which
was observed as a 3-fold and a 1.4-fold increase in the EC50
value of the E167C and R290C mutants (Table 2), respectively,
and a 7.5-fold decrease in the mean current amplitude obtained
with the R290C mutant (Fig. 4, right panel; Table 2).
These data support the functional importance of the
electrostatic interaction of Glu167 and Arg290. The partial
restoration of the function of the R290C mutant obtained
78
Hausmann et al.
TABLE 1
ATP potencies, agonist-induced current responses and derived mutation-induced Gibbs free energy changes and interaction free energies of the wt and
mutant P2X2 receptors
rP2X2
EC50 (95% CI)
EC50mut/EC50wt
nH
1 mM ATP
mM
11.4
6.3
7192
328
17.5
3538
553
889
547
11.4
2663
396
24.7
172
182
5.1
377
295
12.8
139
91
10.2
10.5
729
837
(9.3–14.9)
(5.4–7.6)
(6393–8607)
(288–371)
(14.6–25.6)
(2815–4689)
(491–658)
(841–998)
(509–652)
(8.3–15.5)
(2054–3528)
(299–525)
(17.9–34.1
(137–209)
(126–253)
(3.7–7.1)
(346–412)
(277–314)
(9.8–16.6)
(117–164)
(80–104)
(8.8–12.2)
(7.5–14.1)
(637–839)
(784–901)
n
—
0.7
0.002
0.4
0.5
0.001
0.05
0.1
0.2
1.0
0.04
0.5
0.3
0.1
0.3
1.1
0.2
0.7
1.0
0.7
1.6
—
0.5b
0.05b
0.2b
9
6
6
6
5
4
4
5
4
8
8
7
5
5
5
10
10
11
9
10
10
7
7
7
8
DDG
DDGINT
Imax, mA
—
0.6
624
29
1.5
308
48
77
48
1
234
35
2
15
16
0.5
33
26
1.1
12
8
—
1b
71b
82b
1.5
1.8
2.6
2.6
0.9
1.1
2.5
1.9
2.4
1.5
2.6
2.4
1.0
1.8
1.4
1.3
2.5
2.0
1.3
2.0
1.6
1.4
1.5
2.6
2.6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
0.4
0.2
0.3
0.2
0.4
0.1
0.3
0.3
0.3
0.3
0.3
0.7
0.2
0.3
0.5
0.3
0.3
0.1
0.2
0.3
0.1
0.1
0.5
0.4
0.4
53.9
35.9
0.1
23.4
24.6
0.08
2.6
4.1
9.1
51.0
2.4
24.8
17.3
6.2
14.1
61.1
12.3
36.7
54.9
39.5
84.6
55.7
28.6
3.0
8.4
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6.6
2.1
0.02
1.7
5.1
0.01
0.4
2.3
3.5
3.4
1.9
3.7
2.2
1.9
7.1
5.5
0.7
1.3
4.2
3.6
3.9
4.9
4.8
0.9
2.4
20.34
3.75
1.95
0.25
3.34
2.25
2.53
2.25
20.01
3.17
2.06
0.44
1.58
1.61
20.47
2.04
1.89
0.07
1.45
1.21
0
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
0.05
0.06
0.07
0.02
0.01
0.06
0.09
0.07
0.04
0.01
0.02
0.04
0.02
0.06
0.05
0.09
0.10
0.02
0.04
0.06
21.45 6 0.03
21.33 6 0.08
21.46 6 0.05
20.74 6 0.03
21.11 6 0.03
20.41 6 0.04
0.32 6 0.07
20.31 6 0.04
ATP, adenosine-5’-triphosphate; DDG, calculated Gibbs free energy changes of the mutant; DDGINT, calculated interaction free energy (see Materials and Methods); Imax,
mean current amplitude elicited by 1 mM ATP; n, number of independent measurements; nH, Hill coefficient.
a
Generated using the rP2X2C9.348,430S receptor.
b
Normalized to the rP2X2C9.348,430S receptor.
through the restoration of the positive charge through the
MTSEA reagent indicates that the specific side chain of
arginine itself might also be functionally important. The
marked changes in the ATP-induced current amplitudes of
the R290C mutant treated with MTSEA or MTSES suggest
that the charge in position 290 plays a critical role not only in
ATP binding (Hattori and Gouaux, 2012) but also in channel
gating.
Disulfide Trapping Confirms the Proximity of Residues 167 and 290. To confirm the ∼3 Å spatial proximity
between Glu167 and Arg290 that was predicted by our
homology model (Fig. 1), we substituted Glu167 and Arg290
with cysteines and tested whether these cysteines could be
oxidatively cross-linked. The ATP-induced currents recorded
from the oocytes that expressed the cysteine double mutant
E167C,R290C-rP2X2C9,348,430S were statistically significantly
(P , 0.001) lower after a 120-second exposure to 0.3% H2O2
(Fig. 5). The H2O2-induced inhibition of the ATP-evoked current
could be partially reversed by the subsequent incubation of the
oocytes with the disulfide-reducing agent DTT (Fig. 5). In one
set of experiments with four oocytes (from one batch) that
expressed the E167C,R290C-rP2X2C9,348,430S receptor, the
current amplitude in response to ATP was larger after DTT
application compared with the initial current amplitude before
Fig. 3. Double-mutant cycle analysis of Glu167/Arg290 and
Asp57/Arg275 rP2X2 mutants. The ATP EC50 values of the
indicated rP2X2 receptor constructs were derived from the
concentration-response curves (see Fig. 2B and Table 1).
Significant interaction free energies greater than 60.35
kcal/mol (Schreiber and Fersht, 1995) were obtained with
the Glu167/Arg290 pair of residues but not the As57/Arg275
pair of residues, which are separated by more than 30 Å in
our rP2X2 receptor homology model.
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
wt
E167R
R290E
E167R, R290E
E167K
R290D
E167K, R290D
E167K, R290E
E167R, R290D
E167A
R290A
E167A, R290A
E167D
R290K
E167D, R290K
D57R
R275D
D57R, R275D
D57A
R275A
D57A, R275A
rP2X2C9.348,430S
E167Ca
R290Ca
E167C, R290Ca
Imaxmut/Imaxwt
Salt Bridge Switching in P2X2 Receptor Gating
79
Fig. 4. The effect of the methanethiosulfonate
(MTS) reagents on the maximal currents of the
rP2X2C9,348,430S receptor and the E167C and
R290C mutants. The bars show the relative
changes in the mean current amplitudes 6
S.E.M. of the indicated rP2X2C9,348,430S receptor constructs in response to 1 mM ATP
before and after a 120-second pulse with
the positively and negatively charged MTS
reagents 2-aminoethyl methanethiosulfonate
(MTSEA) and 2-sulfonatoethyl methanethiosulfonate (MTSES), respectively. Consistent
with the proposed electrostatic interaction of
Glu167 and Arg290, the reintroduction of the
negative charge at C167 by MTSES and the
reintroduction of the positive charge at C290
by MTSEA increased the ATP-elicited current
amplitudes.
unchanged EC50 value suggests, but does not prove, that the
disulfide bond formation affected the channel gating rather
than the ATP binding.
We further examined whether the disulfide bond formation
between E167C and R290C occurs in the open state of the
rP2X2C9,348,430S receptor. We therefore tested the ability of
H2O2 to promote cysteine cross-linking in the presence of
3 mM ATP, which is a concentration that approximately
corresponds to the EC95 value of the E167C,R290CrP2X2C9,348,430S receptor (Fig. 5C; Table 1). When H2O2 was
applied to the open E167C, R290C-rP2X2C9,348,430S channel,
only a marginal inhibition of the current amplitude was
observed in the continuous presence of ATP (Fig. 6A). In
addition, subsequent repetitive stimulations with ATP provided no evidence of the current reduction that is indicative
of an inhibitory disulfide bond formation similar to that
observed when H2O2 was applied in the absence of ATP
(Fig. 6A). This finding suggests that the presence of ATP prevents the disulfide bond formation between E167C and R290C
by hindering the contact between the two cysteine residues
either by steric hindrance or by a conformational increase in
the distance between 167C and 290C during ATP-induced
channel gating.
We attempted to discriminate between these two possibilities by testing whether the competitive rP2X2 receptor
antagonist NF770 blocks the ability of H2O2 to promote
disulfide formation in the E167C,R290C-rP2X2C9,348,430S
TABLE 2
ATP potencies and agonist-induced current responses of the indicated P2X2 receptor mutants before and
after MTS treatment
rP2X2
MTS
EC50 (95% CI)
nH
1 mM ATP
rP2X2C9.348,430S
E167C-rP2X2C9.348,430S
R290C-rP2X2C9.348,430S
—
MTSEA
MTSES
—
MTSEA
MTSES
—
MTSEA
MTSES
8.5
8.6
7.9
12.9
39.6
6.6
896
437
1284
(6.4–10.2)
(7.9–9.5)
(6.9–8.7)
(10.9–15.3)
(33.9–46.1)
(5.5–8.1)
(839–981)
(411–473)
(1229–1343)
n
Imax, mA
mM
1.6
1.4
1.6
1.7
1.2
2.3
2.4
1.8
2.3
6
6
6
6
6
6
6
6
6
0.3
0.3
0.3
0.2
0.3
0.3
0.3
0.6
0.2
54.1
56.7
47.6
24.9
19.9
29.0
3.0
8.3
0.4
6
6
6
6
6
6
6
6
6
5.3
4.7
5.8
4.9
1.6
1.8
0.9
1.0
0.1
6
5
5
6
5
5
5
6
6
ATP, adenosine-5’-triphosphate; CI, confidence interval; MTS, methanethiosulfonate; n, number of independent
measurements; nH, Hill coefficient.
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
H2O2 treatment. This is consistent with experiments in which
DTT application increased the ATP-induced current amplitude of the E167C,R290C-rP2X2C9,348,430S receptor without
prior H2O2 addition (data not shown). These data suggest that
the two cysteines can spontaneously form a disulfide bond.
Similarly, variable levels of spontaneous disulfide bond
formation have been previously observed in different batches
of oocytes (Liu et al., 2006). Neither H2O2 nor DTT produced
a statistically significant change in the current amplitudes
(Fig. 5B) obtained with oocytes that expressed the parental
rP2X2C9,348,430S receptor or one of the cysteine single mutants
(E167C or R290C). Overall, these data demonstrate that H2O2
induces a disulfide cross-link between E167C and R290C that
reduces the ATP-gated currents and most likely locks the
E167C,R290C-rP2X2C9,348,430S mutant in a less activatable
state.
To determine whether the reduced current amplitude after
the oxidative disulfide cross-linking of the E167C,R290CrP2X2C9,348,430S mutant is a result of impaired ATP binding
or channel gating, we performed an ATP concentration–
response analysis before and after H2O2 treatment. The
H2O2 treatment did not significantly affect the EC50 value
[850 (95% CI 6983–1034) mM ATP and 1002 (95% CI
772–1301) mM ATP before and after H2O2 treatment, respectively] of the E167C,R290C-rP2X2C9,348,430S mutant, but
reduced the mean current amplitude elicited by 3 mM ATP to
∼67% (Fig. 5C). The reduced current amplitude at an almost
80
Hausmann et al.
concentration–response analysis yielded EC50 values of 850 (95% CI
6983–1034) mM (nH = 2.7 6 0.8, Imax = 8.4 6 0.6 mA, n = 6) and 1002 (95%
CI 772–1301) mM (nH = 2.8 6 0.8, Imax = 5.8 6 0.6 mA, n = 6) for the
nonoxidized (j) and H2O2-oxidized (m) E167C,R290C-rP2X2C9,348,430S
mutants, respectively. The absolute current amplitudes are plotted.
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
Fig. 5. Disulfide trapping of cysteines substituted at Glu167 and Arg290
of the rP2X2 receptor. (A) Original current traces of the E167C,R290CrP2X2C9,348,430S receptor elicited by 1 mM ATP (black bar) before and after
application of H2O2 for 120 seconds (white arrow) or dithiothreitol (DTT)
for 180 seconds (gray arrow). (B) The bars show the change in the mean
current amplitudes 6 S.E.M. (n = 7–12) obtained with the rP2X2C9,348,430S
receptor and its mutants; these were calculated by normalizing the
current amplitudes after exposure to H2O2 or DTT to the current
amplitude before exposure to H2O2 or DTT. The asterisks denote a highly
statistically significant change in the maximal current (P , 0.001) for the
indicated mutant compared with other rP2X2 receptor constructs. No
other pairwise comparisons were statistically significant. (C) The ATP
receptor even if the channel is closed. We have shown that
NF770 fits tightly into the ATP-binding pocket and
interacts strongly with Arg290 through a methoxy group
(Wolf et al., 2011). Thus, if steric hindrance is responsible
for the noninhibitory effect of H2O2 in the presence of ATP,
then NF770, similarly to ATP, should diminish the H2O2induced disulfide cross-linking of E167C and R290C. Upon
coapplication with 3 mM ATP, 10 mM NF770 inhibited the
ATP-induced current amplitude of the E167C,R290CrP2X2C9,348,430S mutant by ∼50%. After NF770 washout,
the magnitude of the current response to ATP was the same
as the current obtained before exposure to NF770 (Fig. 6, B
and C). When 10 mM NF770 was coapplied with H2O2 to the
closed channel, which was followed by the washout of both
compounds, the current response to ATP was diminished to
∼41% (39–44%), which is similar to the extent that was
observed when H2O2 was applied in the absence of NF770
(Fig. 6, B and D).
We also examined the extent of inhibitory disulfide bond
formation when 3 mM ATP was coapplied with 10 mM NF770
and H2O2 (Fig. 6C). The current amplitude remained stable as
long as ATP, NF770, and H2O2 were applied together. After
the washout of the NF770 and H2O2 mix, the ATP-induced
current amplitude was determined to be 20% smaller than
that obtained with the previous H2O2 treatment (Fig. 6, C and
D). The data indicate that the fraction of channels that were
maintained in the closed state by the competitive inhibition of
NF770 was sensitive to the inhibitory disulfide bond formed
between E167C and R290C. The observed 20% reduction in
the current amplitude corresponds to a ∼40% inhibition of
50% of the E167C,R290C-rP2X2C9,348,430S channels that were
maintained in the closed, cross-linkable state by NF770. In
other words, the same ∼40% proportion of closed channels
became oxidatively blocked regardless of whether all or some
of the channels were in the closed conformation. Overall,
there seems to be a proportional relationship between the
ratio of open to closed channels and the extent of channel
inhibition by oxidation. Thus, the NF770 data suggest that
the open channel state, as opposed to steric hindrance by
a ligand, prevents the disulfide bond formation in the
presence of ATP.
To exclude the possibility that ATP sterically blocks access
to the cysteine side chains introduced by the E167C and
R290C mutations, we examined whether the presence of ATP
allows the modification of single cysteine mutants by the
cationic or anionic MTS reagents. MTSES produced the same
small but reproducible increase in the E167C-rP2X2C9,348,430S
receptor-mediated current amplitude in the absence or
presence of 3 mM ATP (Fig. 7, left panel). Also the MTSEAinduced increase in the R290C-rP2X2C9,348,430S receptormediated current amplitude was the same in the absence
and presence of 3 mM ATP (Fig. 7, right panel). Overall, the
data suggest that the reduced level of cross-linking in the
ATP-bound open state results from a conformational increase
Salt Bridge Switching in P2X2 Receptor Gating
81
Fig. 6. Effect of ATP and/or NF770 on the disulfide trapping of the
E167C,R290C-rP2X2C9,348,430S receptor. The typical current traces are
shown, which were monitored using the same protocol as in Fig. 5 except
that (A) 3 mM ATP, (B) 10 mM NF770, or (C) 3 mM ATP + 10 mM NF770
was coapplied with 0.3% H2O2, as indicated. Note that the coapplication of
3 mM ATP with 10 mM NF770 elicited an approximately 50% decrease in
the current amplitude compared with the application of 1 mM ATP. (D)
The bars show the changes in the mean current amplitudes 6 S.E.M. (n =
6-12) of the E167C,R290C-rP2X2C9,348,430S receptor treated with H2O2 in
the absence or presence of ATP or/and NF770, as indicated; these
amplitudes were normalized to the current amplitude before H2O2
treatment. The asterisks indicate whether the respective experimental
condition resulted in a statistically significant difference (*P , 0.05; **P ,
0.01; ***P , 0.001) compared with the treatment with H2O2 alone. n.s.,
not statistically significant.
in the distance between Glu167 and Arg290 upon P2X2
receptor activation.
Discussion
Ionic interactions, which are also designated as salt
bridges, have been identified within and between subunits
and have been shown to be critical for gating, pore opening,
and stability in various ion channels. In the superfamily of
Cys-loop receptors, intrasubunit salt bridges are responsible
for coupling agonist binding to channel gating (Kash et al.,
2003) and for stabilizing the ligand-bound state by restricting
the loop C mobility of GABAA receptors (Venkatachalan and
Czajkowski, 2008). An intersubunit salt bridge at the b/a
interface of the GABAA receptor was determined to stabilize
the ligand-bound state (Laha and Wagner, 2011). In the
monomeric outer-membrane protein A (OmpA) channel in
Escherichia coli, switching between two salt bridges within
the pore defines the closed and open states of the channel
(Moroni and Thiel, 2006; Hong et al., 2006).
In this study, we provide different evidence to support the
existence of a functionally important electrostatic interaction between Glu167 and Arg290 in the ATP-binding pocket of
the closed rP2X2 receptor that is broken upon channel opening: 1) the charge swapping of Glu167 and Arg290 partially
rescued the decrease in the receptor function that was caused
by the single charge-reversal mutations; 2) the recharging of
E167C with the negatively charged MTSES and the recharging of R290C with the positively charged MTSEA partially
rescued the rP2X2 receptor function; 3) the modification of
E167C or R290C with the respective oppositely charged MTS
reagent led to a further decline in the receptor function; 4)
the charge-swap (E167R,R290E-rP2X2) and double-charge
neutralization (E167A,R290A-rP2X2) mutants nonadditively
altered the ATP potency, and the double-mutant cycle analysis revealed significant interaction free energies, which
indicated that Glu167 and Arg290 are energetically coupled;
and 5) the cysteine substitution of Glu167 and Arg290
resulted in oxidative intrasubunit disulfide cross-linking between these two residues in the closed but not the open state
of the P2X2 receptor channel. Based on these data along with
the partial blocking of the ATP-induced current responses
after disulfide formation, we infer that the salt bridge between Glu167 and Arg290 serves to stabilize the closed state
of the receptor.
We consider it unlikely that the Glu167/Arg290 salt bridge
is buried and hence water inaccessible for several reasons:
1) according to the rP2X2 homology model, the Glu167 and
Arg290 residues line the surface of the ATP-binding pocket,
and 2) both the E167C and the R290C mutants react readily
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
Fig. 7. The effects of the methanethiosulfonate (MTS) reagents on the
E167C- and R290C-rP2X2C9,348,430S mutant currents in the presence of
ATP. The bars show the relative changes in the mean ATP-induced
current amplitudes after treatment with 2-aminoethyl methanethiosulfonate (MTSEA) or 2-sulfonatoethyl methanethiosulfonate (MTSES) for 120
seconds in presence or absence of 3 mM ATP.
82
Hausmann et al.
electrostatic interaction against a complicated background
(Horovitz et al., 1990; Serrano et al., 1990). Therefore, the
unchanged ATP potency of the E167A mutant does not argue
against a stabilization of the closed state of the rP2X2
receptor by the Glu167/Arg290 salt bridge.
The observed disulfide bond formation between E167C and
R290C strongly indicates that the distance between these two
residues is less than 4.6 Å, which is the maximum distance
between the Cb-carbons of cysteines that can be linked by
a disulfide bond (Sowdhamini et al., 1989; Careaga and Falke,
1992a, 1992b). We propose that a gating movement of the
E167C and R290C residues upon channel activation by ATP,
and not steric hindrance by bound ATP, accounts for the
impaired disulfide bond formation between these two residues
in the presence of ATP for the following reasons: 1) the
oxidative disulfide bridge formation did not affect the ATP
potency but reduced the ATP efficacy, which suggests that the
disulfide bond interfered with channel gating and not ATP
binding; 2) the competitive antagonist NF770, which interacts
with several basic residues in the ATP-binding pocket in
a manner similar to the phosphate oxygen atoms of ATP (Wolf
et al., 2011), did not affect the oxidative disulfide formation;
and 3) ATP did not prevent the reaction of E167C and R290C
with the MTS reagents, which suggests that the bound ATP
does not block the access of the introduced cysteines to the
thiol-reactive reagents. Overall, these data suggest that the
reduced level of cross-linking in the ATP-bound open-channel
state results from a spatial rearrangement of the Glu167 and
Arg290 residues in the open-channel state, which is a result of
the increased separation of these residues during channel
gating.
A ∼0.5- Å increase in the side chain distance between the
carboxylate carbon of Glu163 and the guanidinium carbon of
Arg298 upon the closed-to-open transition is also indicated by
a comparison of the apo-closed-state (4DW0) and the ATP-bound
Fig. 8. Comparison of the closed- and open-state rP2X2 receptor homology models. Selected details of the interior of the intrasubunit ATP-binding
pocket of the apo-closed state (left panel) and the ATP-bound open state (right panel) of the homology-modeled rP2X2 receptor (viewed from the side,
parallel to the membrane plane) are presented. The models are based on the closed (PDB entry 3H9V) and open (PDB entry 4DW1) zfP2X4 X-ray
structures. The head domain and the C-terminal base of the left “flipper” are located in the upper right and lower left corner, respectively. The
polypeptide backbones of the two adjacent rP2X2 subunits that contribute to the ATP-binding site are colored in orange or pink, and selected residues
are depicted as sticks. Specific atoms of the amino acid side chains and ATP are colored (carbons are light gray or yellow, oxygens are red, nitrogens are
blue, hydrogens are dark gray, and phosphates are pink). A comparison of the closed- and open-state models indicates substantial movement of the
protein backbone containing the Glu167 and Arg290 residues, which includes a marked rearrangement of their side chains. In the closed state, Arg290 is
found to be mainly in an ionic interaction with Glu167 (dashed light blue lines marked by a gray arrowhead, left panel). In contrast, in the ATP-bound
open state, Arg290 is observed mainly in a strong ionic interaction with a g-phosphate oxygen of ATP (dashed light blue lines marked by a gray arrow,
right panel). The thickness of the center of the dashed blue lines indicates the strength of the ionic interaction, which was assessed using the
MOE2008.10 program.
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
with the charged MTS reagents, which strongly argues
against buried positions of the engineered cysteine residues.
The hydrophobic solubility of a cysteine residue is sufficiently high to not prevent the burial of the engineered
cysteines in the protein core (Nagano et al., 1999; Hessa
et al., 2005). In addition, both cysteine residues in the
E167C,R290C-rP2X2C9,348,430S double mutant were readily
accessible for oxidative cross-linking, and the disulfide bond
that was formed between them could be readily cleaved by
chemical reduction. The relatively low DDGINT of less than
|2| kcal/mol is also consistent with a localization of the
Glu167/Arg290 salt bridge in a hydrophilic environment. Salt
bridges within water-accessible, hydrophilic environments
are generally known to be of lower strength than ionic interactions within buried hydrophobic regions of the protein
interior (Kumar and Nussinov, 1999). Altogether, the negative DDGINT deduced from the mutant cycle analysis can be
most readily assigned to a surface-localized, water-exposed
salt bridge that stabilizes the closed state of the rP2X2
receptor.
We note that the breaking of a closed-state, stabilizing salt
bridge by charge neutralization may be expected to favor
channel opening and thus increase the agonist potency. In the
large data set analyzed in this study, the E167A mutation did
not significantly change the ATP potency of the rP2X2
receptor in either direction. However, breaking a salt bridge
by simply mutating one of the pair of charges most likely
changes a number of interactions within the protein that
contribute to its stability (Serrano et al., 1990). In the case of
the rP2X2 receptor, the charge neutralization of Glu167
releases Arg290 from its coupling to the residue at position
167. The uncoupled Arg290 may undergo compensatory interactions with other acidic residues that may cancel out the
destabilizing effect of the E167A mutation. Only the doublemutant cycle allows one to isolate the energy of the
Salt Bridge Switching in P2X2 Receptor Gating
Authorship Contributions
Participated in research design: Hausmann, Schmalzing.
Conducted experiments: Hausmann, Günther, Kless, Kuhlmann,
Markwardt.
Contributed new reagents or analytic tools: Kassack.
Performed data analysis: Hausmann, Günther, Kless, Kuhlmann,
Bahrenberg, Markwardt.
Wrote or contributed to the writing of the manuscript: Hausmann,
Schmalzing.
References
Allsopp RC, El Ajouz S, Schmid R, and Evans RJ (2011) Cysteine scanning mutagenesis (residues Glu52-Gly96) of the human P2X1 receptor for ATP: mapping
agonist binding and channel gating. J Biol Chem 286:29207–29217.
Aschrafi A, Sadtler S, Niculescu C, Rettinger J, and Schmalzing G (2004) Trimeric
architecture of homomeric P2X2 and heteromeric P2X1+2 receptor subtypes. J Mol
Biol 342:333–343.
Cao L, Broomhead HE, Young MT, and North RA (2009) Polar residues in the second
transmembrane domain of the rat P2X2 receptor that affect spontaneous gating,
unitary conductance, and rectification. J Neurosci 29:14257–14264.
Careaga CL and Falke JJ (1992a) Structure and dynamics of Escherichia coli chemosensory receptors. Engineered sulfhydryl studies. Biophys J 62:209–216, discussion 217–219.
Careaga CL and Falke JJ (1992b) Thermal motions of surface alpha-helices in the Dgalactose chemosensory receptor. Detection by disulfide trapping. J Mol Biol 226:
1219–1235.
Carter PJ, Winter G, Wilkinson AJ, and Fersht AR (1984) The use of double mutants
to detect structural changes in the active site of the tyrosyl-tRNA synthetase
(Bacillus stearothermophilus). Cell 38:835–840.
Du J, Dong H, and Zhou HX (2012) Gating mechanism of a P2X4 receptor developed
from normal mode analysis and molecular dynamics simulations. Proc Natl Acad
Sci USA 109:4140–4145.
Ennion S, Hagan S, and Evans RJ (2000) The role of positively charged amino acids
in ATP recognition by human P2X(1) receptors. J Biol Chem 275:29361–29367.
Faiman GA and Horovitz A (1996) On the choice of reference mutant states in the
application of the double-mutant cycle method. Protein Eng 9:315–316.
Fallah G, Römer T, Detro-Dassen S, Braam U, Markwardt F, and Schmalzing G
(2011) TMEM16A(a)/anoctamin-1 shares a homodimeric architecture with CLC
chloride channels. Mol Cell Proteomics 10:M110.004697.
Fischer W, Zadori Z, Kullnick Y, Gröger-Arndt H, Franke H, Wirkner K, Illes P,
and Mager PP (2007) Conserved lysin and arginin residues in the extracellular
loop of P2X(3) receptors are involved in agonist binding. Eur J Pharmacol 576:
7–17.
Franklin C, Braam U, Eisele T, Schmalzing G, and Hausmann R (2007) Lack of
evidence for direct phosphorylation of recombinantly expressed P2X(2) and P2X (3)
receptors by protein kinase C. Purinergic Signal 3:377–388.
Guerlet G, Taly A, Prado de Carvalho L, Martz A, Jiang R, Specht A, Le Novère N,
and Grutter T (2008) Comparative models of P2X2 receptor support inter-subunit
ATP-binding sites. Biochem Biophys Res Commun 375:405–409.
Hattori M and Gouaux E (2012) Molecular mechanism of ATP binding and ion
channel activation in P2X receptors. Nature 485:207–212.
Hausmann R, Bodnar M, Woltersdorf R, Wang H, Fuchs M, Messemer N, Qin Y,
Günther J, Riedel T, and Grohmann M et al. (2012) ATP binding site mutagenesis
reveals different subunit stoichiometry of functional P2X2/3 and P2X2/6 receptors.
J Biol Chem 287:13930–13943.
Hausmann R, Rettinger J, Gerevich Z, Meis S, Kassack MU, Illes P, Lambrecht G,
and Schmalzing G (2006) The suramin analog 4,49,40,49-(carbonylbis(imino-5,1,3benzenetriylbis (carbonylimino)))tetra-kis-benzenesulfonic acid (NF110) potently
blocks P2X3 receptors: subtype selectivity is determined by location of sulfonic acid
groups. Mol Pharmacol 69:2058–2067.
Hessa T, Kim H, Bihlmaier K, Lundin C, Boekel J, Andersson H, Nilsson I, White
SH, and von Heijne G (2005) Recognition of transmembrane helices by the endoplasmic reticulum translocon. Nature 433:377–381.
Hidalgo P and MacKinnon R (1995) Revealing the architecture of a K+ channel pore
through mutant cycles with a peptide inhibitor. Science 268:307–310.
Hong H, Szabo G, and Tamm LK (2006) Electrostatic couplings in OmpA ion-channel
gating suggest a mechanism for pore opening. Nat Chem Biol 2:627–635.
Horovitz A (1996) Double-mutant cycles: a powerful tool for analyzing protein
structure and function. Fold Des 1:R121–R126.
Horovitz A, Serrano L, Avron B, Bycroft M, and Fersht AR (1990) Strength and cooperativity of contributions of surface salt bridges to protein stability. J Mol Biol
216:1031–1044.
Jiang R, Martz A, Gonin S, Taly A, de Carvalho LP, and Grutter T (2010) A putative
extracellular salt bridge at the subunit interface contributes to the ion channel
function of the ATP-gated P2X2 receptor. J Biol Chem 285:15805–15815.
Jiang R, Taly A, Lemoine D, Martz A, Cunrath O, and Grutter T (2012) Tightening of
the ATP-binding sites induces the opening of P2X receptor channels. EMBO J 31:
2134–2143.
Kaczmarek-Hájek K, Lörinczi E, Hausmann R, and Nicke A (2012) Molecular and
functional properties of P2X receptors—recent progress and persisting challenges.
Purinergic Signal 8:375–417.
Kash TL, Jenkins A, Kelley JC, Trudell JR, and Harrison NL (2003) Coupling
of agonist binding to channel gating in the GABA(A) receptor. Nature 421:272–275.
Kawate T, Michel JC, Birdsong WT, and Gouaux E (2009) Crystal structure of the
ATP-gated P2X(4) ion channel in the closed state. Nature 460:592–598.
Kawate T, Robertson JL, Li M, Silberberg SD, and Swartz KJ (2011) Ion access
pathway to the transmembrane pore in P2X receptor channels. J Gen Physiol 137:
579–590.
Kracun S, Chaptal V, Abramson J, and Khakh BS (2010) Gated access to the pore of
a P2X receptor: structural implications for closed-open transitions. J Biol Chem
285:10110–10121.
Kumar S and Nussinov R (1999) Salt bridge stability in monomeric proteins. J Mol
Biol 293:1241–1255.
Laha KT and Wagner DA (2011) A state-dependent salt-bridge interaction exists
across the b/a intersubunit interface of the GABAA receptor. Mol Pharmacol 79:
662–671.
Li M, Chang TH, Silberberg SD, and Swartz KJ (2008) Gating the pore of P2X
receptor channels. Nat Neurosci 11:883–887.
Li M, Kawate T, Silberberg SD, and Swartz KJ (2010) Pore-opening mechanism in
trimeric P2X receptor channels. Nat Commun 1:44.
Liu X, Alexander C, Serrano J, Borg E, and Dawson DC (2006) Variable reactivity of
an engineered cysteine at position 338 in cystic fibrosis transmembrane conductance regulator reflects different chemical states of the thiol. J Biol Chem 281:
8275–8285.
Meis S, Hamacher A, and Hongwiset D et al. (2010) NF546 [4,49-(carbonylbis(imino3,1-phenylene-carbonylimino-3,1-(4-methyl-phenylene)-carbonylimino))-bis(1,3-xylene-alpha,alpha9-diphosphonic acid) tetrasodium salt] is a non-nucleotide P2Y11
agonist and stimulates release of interleukin-8 from human monocyte-derived
dendritic cells. J Pharmacol Exp Ther 332:238–247.
Moroni A and Thiel G (2006) Flip-flopping salt bridges gate an ion channel. Nat Chem
Biol 2:572–573.
Nagano N, Ota M, and Nishikawa K (1999) Strong hydrophobic nature of cysteine
residues in proteins. FEBS Lett 458:69–71.
Nicke A, Bäumert HG, Rettinger J, Eichele A, Lambrecht G, Mutschler E,
and Schmalzing G (1998) P2X1 and P2X3 receptors form stable trimers: a novel
structural motif of ligand-gated ion channels. EMBO J 17:3016–3028.
Nicke A, Rettinger J, Büttner C, Eichele A, Lambrecht G, and Schmalzing G (1999)
Evolving view of quaternary structures of ligand-gated ion channels. Prog Brain
Res 120:61–80.
North RA (2002) Molecular physiology of P2X receptors. Physiol Rev 82:1013–1067.
Rettinger J and Schmalzing G (2003) Activation and desensitization of the recombinant
P2X1 receptor at nanomolar ATP concentrations. J Gen Physiol 121:451–461.
Roberts JA, Allsopp RC, El Ajouz S, Vial C, Schmid R, Young MT, and Evans RJ
(2012) Agonist binding evokes extensive conformational changes in the
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
open-state (4DW1) X-ray structures of the zfP2X4 receptor.
The Glu163 and Ar298 residues of the zfP2X4 receptor
correspond to the Glu167 and Arg290 residues of the rP2X2
receptor, respectively. Moreover, molecular dynamics simulations suggest that the b7-strand and the b-sheet of the b13- to
b14-strands, which contain Glu163 and Arg298, respectively,
undergo significant movement during channel gating (Du
et al., 2012).
The comparison of the rP2X2 receptor homology models,
which were based on the closed- and open-state crystal
structures of the zfP2X4 receptor, reveal substantial movement of the protein backbone containing Glu167 and Arg290
as well as a marked rearrangement of their side chains in the
closed-to-open state transition (Fig. 8). The closed-state
rP2X2 model shows that Arg290 is mainly in ionic contact
with Glu167 (dashed blue lines marked by a gray arrowhead;
Fig. 8, left panel). In contrast, in the ATP-bound open-state
model of the rP2X2 receptor, Arg290 is mainly detected to
form a strong ionic interaction with a g-phosphate oxygen
of ATP (dashed blue lines marked by a gray arrow; Fig. 8,
right panel). Thus, because of the spatial rearrangement, the
electrostatic coupling between Glu167 and Arg290 is released,
which results in a new ionic interaction between Arg290 and
ATP. The importance of Arg290 in ATP binding through the
coordination of a g-phosphate oxygen of ATP is indicated in
the P2X2 receptor model and supported by the site-directed
mutagenesis of Arg290. The charge neutralization at position
290 resulted in a marked decrease in the ATP potency; this
effect was further decreased by charge reversal mutants with
glutamic acid or aspartic acid. Thus, our results imply that
Arg290 has a dual state-dependent role in the P2X2 receptor
function: the Arg290/Glu167 salt bridge stabilizes the closed
state of the receptor, whereas the Arg290/ATP interaction is
of crucial importance for the coordination of ATP in the open
ATP-bound state.
83
84
Hausmann et al.
extracellular domain of the ATP-gated human P2X1 receptor ion channel. Proc
Natl Acad Sci USA 109:4663–4667.
Roberts JA, Digby HR, Kara M, El Ajouz S, Sutcliffe MJ, and Evans RJ (2008)
Cysteine substitution mutagenesis and the effects of methanethiosulfonate reagents at P2X2 and P2X4 receptors support a core common mode of ATP action at
P2X receptors. J Biol Chem 283:20126–20136.
Roberts JA and Evans RJ (2004) ATP binding at human P2X1 receptors. Contribution of aromatic and basic amino acids revealed using mutagenesis and partial
agonists. J Biol Chem 279:9043–9055.
Samways DS, Khakh BS, Dutertre S, and Egan TM (2011) Preferential use
of unobstructed lateral portals as the access route to the pore of human ATP-gated
ion channels (P2X receptors). Proc Natl Acad Sci USA 108:13800–13805.
Schmalzing G, Gloor S, Omay H, Kröner S, Appelhans H, and Schwarz W (1991) Upregulation of sodium pump activity in Xenopus laevis oocytes by expression of
heterologous b 1 subunits of the sodium pump. Biochem J 279:329–336.
Schreiber G and Fersht AR (1995) Energetics of protein-protein interactions: analysis
of the barnase-barstar interface by single mutations and double mutant cycles.
J Mol Biol 248:478–486.
Serrano L, Horovitz A, Avron B, Bycroft M, and Fersht AR (1990) Estimating the
contribution of engineered surface electrostatic interactions to protein stability by
using double-mutant cycles. Biochemistry 29:9343–9352.
Sowdhamini R, Srinivasan N, Shoichet B, Santi DV, Ramakrishnan C, and Balaram
P (1989) Stereochemical modeling of disulfide bridges. Criteria for introduction
into proteins by site-directed mutagenesis. Protein Eng 3:95–103.
Venkatachalan SP and Czajkowski C (2008) A conserved salt bridge critical for
GABA(A) receptor function and loop C dynamics. Proc Natl Acad Sci USA 105:
13604–13609.
Wolf C, Rosefort C, Fallah G, Kassack MU, Hamacher A, Bodnar M, Wang H, Illes P,
Kless A, and Bahrenberg G et al. (2011) Molecular determinants of potent P2X2
antagonism identified by functional analysis, mutagenesis, and homology docking.
Mol Pharmacol 79:649–661.
Address correspondence to: Ralf Hausmann, Department of Molecular
Pharmacology, RWTH Aachen University, Wendlingweg 2, D-52074 Aachen,
Germany. E-mail address: [email protected]
Downloaded from molpharm.aspetjournals.org at ASPET Journals on June 16, 2017
Molecular Pharmacology
Salt bridge switching from Arg290/Glu167 to Arg290/ATP promotes the closedto-open transition of the P2X2 receptor
Ralf Hausmann, Janka Günther, Achim Kless, Daniel Kuhlmann, Matthias U. Kassack, Gregor
Bahrenberg, Fritz Markwardt, and Günther Schmalzing
zfP2X4
rP2X2
36
zfP2X4
rP2X2
95
zfP2X4
rP2X2
154
zfP2X4
rP2X2
213
zfP2X4
rP2X2
273
zfP2X4
rP2X2
333
RFTQALVIAYVIGYVCVYNKGYQDTDT-VLSSVTTKVKGIALTNTSELGERIWDVADYII
RMVQLLILLYFVWYVFIVQKSYQDSETGPESSIITKVKGI--TM-SE--DKVWDVEEYVK
34
PPQEDGSFFVLTNMIITTNQTQSKCAENP-TPASTCTSHRDCKRGFNDARGDGVRTGRCV
PPEGGSVVSIITRIEVTPSQTLGTCPESMRVHSSTCHSDDDCIAGQLDMQGNGIRTGHCV
89
SY-SASVKTCEVLSWCPLEKIVDPPNPPLLADAENFTVLIKNNIRYPKFNFNKRNILPNI
PYYHGDSKTCEVSAWCPVEDGTSD-NHFLGKMAPNFTILIKNSIHYPKFKFSKGNIASQ-
149
NSSYLTHCVFSRKTDPDCPIFRLGDIVGEAEEDFQIMAVRGGVMGVQIRWDCDLDMPQSW
KSDYLKHCTFDQDSDPYCPIFRLGFIVEKAGENFTELAHKGGVIGVIINWNCDLDLSESE
207
CVPRYTFRRLDNKDPDNNVAPGYNFRFAKYYKNSDGTETRTLIKGYGIRFDVMVFGQAGK
CNPKYSFRRLDPK--YDPASSGYNFRFAKYYKINGTTTTRTLIKAYGIRIDVIVHGQAGK
267
FNIIPTLLNIGAGLALLGLVNVICDWI359
FSLIPTIINLATALTSIGVGSFLCDWI351
325
Supplemental Figure 1. Primary sequence alignment between the template (zfP2X4) and the
model (rP2X2). Shown is the sequence alignment of the zebrafish P2X4 subunit (zfP2X4, UniProt
entry Q6NYR1) and the rat P2X2 subunit (rP2X2, UniProt entry P49653) used for the generation of
the apo-closed state and ATP-bound open state rP2X2 receptor homology models. The sequence
alignment was performed using a modified version of the Needleman-Wunsch algorithm
(Needleman and Wunsch, 1970). In this approach, the alignments are computed by optimizing a
function based on residue similarity scores. The function uses the BLOSUM62 amino acid
substitution matrix (Henikoff and Henikoff, 1992) and gap penalties and was constrained, in this
case, by the 10 conserved extracellular cysteine residues (indicated in yellow), which were adjusted
and fixed manually.
References
Henikoff S and Henikoff J G (1992) Amino Acid Substitution Matrices From Protein Blocks. Proc
Natl Acad Sci U S A 89: 10915-10919.
Needleman SB and Wunsch C D (1970) A General Method Applicable to the Search for
Similarities in the Amino Acid Sequence of Two Proteins. J Mol Biol 48: 443-453.